Causal forest method reveals key factors in treatment effects
Explaining Heterogeneous Treatment Effects with Causal Forests and the Hoeffding Functional Decomposition
Machine Learning
Summary
Estimating how different people respond to treatments is often done using complex machine learning methods that are hard to understand. The authors created CausalTreeHFD, a way to break down these complex results into simple parts that show which factors and interactions affect the treatment results most. Their method helps explain why treatments work differently for different groups by identifying important variables. Tests on both fake and real data show that their approach pinpoints relevant factors accurately.
What this means in practice
- •For healthcare data scientists: Identify which patient features most influence different treatment outcomes to improve personalized medicine strategies.
- •For marketing analysts: Detect key customer attributes and their interactions driving response differences to marketing interventions for targeted campaigns.
Authors
Gaëtan Patinier, Clément Bénard, Margaux Törnqvist
Abstract
The estimation of heterogeneous treatment effects is a central objective of causal inference. However, state-of-the-art algorithms to solve this task are based on machine learning principles, and thus behave as black boxes. Consequently, the identification of covariates driving heterogeneity remains a challenging and open problem. To address this limitation, we introduce CausalTreeHFD, an interpretable extension of causal forests, based on the Hoeffding functional decomposition adapted to tree ensembles. Our approach decomposes predictions of conditional average treatment effects (CATE) into contributions from individual covariates and pairwise interactions, providing a direct characterization of the factors driving treatment effect heterogeneity. We establish the consistency of CausalTreeHFD, which converges towards the Hoeffding decomposition of the underlying CATE function, under standard assumptions for the analysis of causal forests. Experiments on simulated and real datasets show that CausalTreeHFD accurately identifies relevant covariates and interactions, and precisely recovers their contributions to heterogeneous treatment effects.